Two Worlds (Shared) Runtime Skeleton v1

This repo provides a deterministic reducer (state machine) skeleton for applying the events emitted by the DSL ruleset.

Scope:

• Event type definitions
• Core game state representation (money, debt, positions, laps, jail/rehab, city ownership/houses, deck/hand)
• applyEvent(...) reducer with invariants and guardrails
• Minimal "command input" / "environment input" typing
• Smoke demo that applies a few events

Not included:

• DSL parser / evaluator (rulescript engine)
• Networking/UI

Documentation

• See docs/ko/ and docs/en/.

Simulation

• src/sim/ provides simulation + metrics skeleton.
• Plug your DSL RulesEngine into RulesEngine.tick(...).

RulesEngine Adapter

This repo expects a fact-based RulesEngine (tickFacts). See src/integration/rulesEngineAdapter.ts for a precompiled-IR adapter skeleton (PrecompiledRulesEngine).

Built-in minimal RulescriptRuntime

A minimal in-process runtime is provided at src/rulescriptRuntime/. It can evaluate precompiled RuleIR against facts.

Usage:

• Compile ruleset to RuleIR[] offline (precompile step).
• Create new PrecompiledRulesEngine({ irJson, runtime: rulescriptRuntime }).

Compile ruleset to precompiled IR

Run:

• npm run build
• npm run compile:ruleset

This produces ruleset_precompiled/shared_ruleset.ir.json.

Then you can load it and run with:

import { PrecompiledRulesEngine } from "./integration/rulesEngineAdapter.js";
import { rulescriptRuntime } from "./rulescriptRuntime/index.js";

const ir = /* load JSON */;
const engine = new PrecompiledRulesEngine({ irJson: ir, runtime: rulescriptRuntime });